Professional Documents
Culture Documents
Effect of Interaction Type On The Characteristics of Pet-Directed Speech in Female Dog Owners
Effect of Interaction Type On The Characteristics of Pet-Directed Speech in Female Dog Owners
DOI 10.1007/s10071-017-1077-7
ORIGINAL PAPER
Received: 18 June 2016 / Revised: 23 January 2017 / Accepted: 31 January 2017 / Published online: 11 February 2017
Ó Springer-Verlag Berlin Heidelberg 2017
Abstract Recent studies focusing on the interspecific nature of the interaction with their dog, suggesting that the
communicative interactions between humans and dogs intended function of these vocal utterances remains to
show that owners use a special speech register when provide dogs with information about their intentions and
addressing their dog. This register, called pet-directed emotions.
speech (PDS), has prosodic and syntactic features similar
to that of infant-directed speech (IDS). While IDS prosody Keywords Pet-directed speech Human–dog relationship
is known to vary according to the context of the commu- Vocalizations Interaction context
nication with babies, we still know little about the way
owners adjust acoustic and verbal PDS features according
to the type of interaction with their dog. The aim of the Introduction
study was therefore to explore whether the characteristics
of women’s speech depend on the nature of interaction Humans and dogs share a long history, as dogs are believed
with their dog. We recorded 34 adult women interacting to be the first species to have been domesticated, approx-
with their dog in four conditions: before a brief separation, imately 32 thousand years ago (Thalmann et al. 2013).
after reuniting, during play and while giving commands. According to several surveys, dogs are now often consid-
Our results show that before separation women used a low ered as ‘‘part of the family’’ by their owners who see
pitch, few modulations, high intensity variations and very themselves as ‘‘pet parents’’ rather than as ‘‘owners’’
few affective sentences. In contrast, the reunion interac- (Taylor 2006; Del Monte Foods 2011) and provide inter-
tions were characterized by a very high pitch, few imper- specific nurturing and protective behaviour towards their
atives and a high frequency of affectionate nicknames. pet dogs (Fogle 1992; Askew 1996; Archer 1997; Archer
During play, women used mainly questions and attention- and Monton 2011). Gibson et al. (2014) compare the dogs’
getting devices. Finally when commanding, women mainly integration into human families to a ‘‘child-like immer-
used imperatives as well as attention-getting devices. Thus, sion’’. The affective bond between owners and dogs mir-
like mothers using IDS, female owners adapt the verbal as rors the human parents–infant bond and may have a
well as the non-verbal characteristics of their PDS to the common biological basis. For instance, both owners and
dogs experience an important secretion of oxytocin after a
brief period of cuddling or after sharing a mutual gaze
& Sarah Jeannin (Odendaal and Meintjes 2003; Nagasawa et al.
s.jeannin@u-paris10.fr 2009, 2015). Moreover, owners speak to their dogs using
1
pet-directed speech (PDS), a register that strongly resem-
Laboratoire Ethologie, Cognition, Développement (LECD-
bles the infant-directed speech (IDS) used by humans when
EA3456), Univ Paris Nanterre (UPL), 92000 Nanterre,
France talking to infants (Fernald and Simon 1984; Trainor et al.
2 2000; Prato-Previde et al. 2006). The specific prosody of
UMR 7179, CNRS/MNHN, Ecole Nationale Vétérinaire
d’Alfort (ENVA), 7 Avenue du Général de Gaulle, both PDS and IDS is assumed to draw the addressee’s
94700 Maisons-Alfort, France
123
500 Anim Cogn (2017) 20:499–509
attention, as well as to express friendliness and affection Based on previous studies (Fernald 1989; Trainor et al.
(Trainor et al. 2000; Mitchell 2001). 2000) which show that the specific features of IDS are used
PDS and IDS share syntactic and prosodic aspects that to transmit emotional information and that different types
are distinct from adult-directed speech (ADS): higher of IDS are used in different contexts (Fernald 1989; Trai-
fundamental frequency (also called pitch), greater pitch nor et al. 2000), we predicted that female owners’ PDS
range, more exaggerated vowel contrast, shorter phrases, verbal and non-verbal features would be primarily shaped
simpler grammar, repetitions and slower speech rate by the type of interaction. In the separation condition, we
(Hirsh-Pasek and Treiman 1982; Burnham et al. predicted that women would either (1) reassure their dog
1998, 2002). Authors found that IDS contains hyperartic- before leaving, explaining what was going to happen using
ulated vowels (Burnham et al. 2002); this characteristic mostly declarative sentence types and a soothing tone
appears to be related to the audience’s actual or expected characterized by a low pitch and falling contours like IDS
linguistic competence and may serve a didactic function. in a soothing context (Fernald 1989; Trainor et al. 2000) or
Xu et al. (2013) show that the degree of vowel hyperar- (2) prevent the dog from coming with them, using a low
ticulation increases from ADS and PDS, to parrot-directed pitch and flat contours with mostly imperative sentence
speech, and then to IDS. Nevertheless, other research on types, like IDS in a prohibition context (Fernald 1989;
puppy-directed speech shows that both IDS and puppy- Trainor et al. 2000). In contrast, in the reunion condition,
directed speech are more hyperarticulated than ADS which considered as one of the most positive human–dog inter-
raises the possibility that IDS hyperarticulation may be actions (Rehn et al. 2013), we expected that women would
more related to emotional expressiveness than to a desire to speak with a high pitch and increased pitch variation and
teach language (Kim et al. 2006). With regard to hyper- predominantly use affective sentence types. Regarding the
articulation, the comparison between PDS and puppy-di- play context, research investigating at human vocalizations
rected speech remains to be done. to dogs found them to be highly repetitive and imperative
The prosody of IDS is known to change according to the suggesting that owners were attempting to ‘‘control the
context of the interaction with the child (Newport et al. dog’’, i.e. to avoid unwanted behaviours (Mitchell and
1977; Fernald 1989; Papoušek et al. 1990, 1991; Trainor Edmonson 1999). A recent study indicates there are dif-
et al. 1997, 2000). Similarly, Pongrácz et al. (2001) indi- ferent types of play which differ in their form and affect
cated that human–dog acoustic communication could be content and that accordingly dog owners’ vocalizations
highly situation-dependent. However, whether speakers vary in their positive and neutral affect (Horowitz and
adjust the acoustic and verbal features according to the type Hecht 2016). We thus expected women to speak with
of interaction when speaking to their dog has not been mostly imperative sentence types, using few declaratives
systematically investigated. and questions, while prosodic features were expected to
The aim of this study was to explore whether variations vary depending on what type of play owner and dog are
appear within PDS according to the type of interaction involved in. We predicted that both during play and when
human and dog are involved in. Towards these ends, we commanding, women would use a high frequency of rep-
focused on women–dog interaction because female owners, etitions with mainly verb and attention-getting devices as
as previously mentioned, are more likely to use PDS than these two conditions imply a succession of actions, and
male owners (Mitchell 2004; Prato-Previde et al. 2006). thus the necessity to get and maintain the dog’s attention
We investigated both the acoustic and linguistic features of towards the activity (Rogers et al. 1993). Finally in the
female owners’ speech as these two dimensions have been command condition, we expected women to use a ‘‘com-
shown to be intimately related (Syrdal and Kim 2008). To manding tone’’ characterized by a low pitch and flat con-
do so, we analysed the speech of female owners addressing tours with mostly imperative sentence types. This
their dog in four different types of interaction: (a) before a imperative style should be different from imperatives in
brief separation, referred below as ‘‘separation’’; (b) after a play which are assumed to function to encourage the dog to
reunion, referred below as ‘‘reunion’’; (c) during play, continue playing (Mitchell 2001).
referred below as ‘‘play’’; and (d) when giving commands, Finally, some studies suggest that people who are not
referred below as ‘‘commands’’. Our experimental condi- parents tend to see their dog as a family member signifi-
tions are based on the Strange Situation test originally set cantly more than people who have children (Berryman
up by Ainsworth and Wittig (1969) and designed to et al. 1984; Albert and Bulcroft 1988; Kidd and Kidd 1989;
investigate human parent–infant attachment (Bowlby Poresky and Daniels 1998; Taylor 2006). Consequently, we
1969). The Strange Situation procedure was adapted for predicted that owners who do not have children would
studying affectional bonds of dogs towards their owners speak with enhanced PDS prosodic features compared to
(Topál et al. 1998; Prato-Previde et al. 2003). those who have children.
123
Anim Cogn (2017) 20:499–509 501
123
502 Anim Cogn (2017) 20:499–509
Fig. 1 Four conditions of interaction between the owner and her dog: a separation, b reunion, c play and d commands
2. Reunion condition the owner and the experimenter form a unique audio sequence per condition, using
came back reentered the room and the owner was free Audacity software (2.0.3). In the ‘‘commands’’ condition,
to interact with her dog as though she had just returned in order to focus specifically on the acoustic properties of
from work or shopping. The interaction lasted an the commands, we excluded the frequent vocal congratu-
average of 30 s until the owner stopped the interaction lations that came after the dog obeyed a command.
with her dog and looked at the experimenter or asked We performed acoustic analyses using a script in Praat
her about the next step. software (5.3.50) on the following measures: (a) mean F0:
3. Play condition The owner was asked to play with her calculated as the average fundamental frequency over the
dog for 1 min. There were no specific rules regarding duration of the signal and (b) F0CV: the coefficient of
how the owner could play with the dog. Different toys variation of F0 over the duration of the signal, estimated as
were available in the room and offered to the owners the standard deviation of F0 divided by mean F0; it is a
like ropes, tennis balls, cuddly toys, etc. Owners could measure of the intonation (F0 contour variations). We also
also use their own toys. calculated (c) IntCV: the coefficient of variation of the
4. Commands condition The owner was asked to give intensity contour (intensity corresponds to the energy in the
commands for 1 min, whether the dog obeyed or not. sound), as modulation of fundamental frequency has been
The owner was not asked to give specific commands, shown to inevitably co-vary with distinctive patterns of
but to interact as usual. It was, however, specified that intensity (Sokol et al. 2005).
she could repeat the same command as many times as
she wanted. The owner was allowed to reward the dog
Verbal analysis
by voice (congratulations) or with treats at her
disposal.
Owners’ speeches utterances were independently tran-
scribed by two investigators. We also noted non-verbal
Data analysis vocal sounds such as whistles. Lapel microphones enabled
a high audio recording quality, even when owners whis-
Acoustic analysis pered. Agreement in word and non-word identifications
between the two investigators was 95%.
During each condition, women’s utterances were generally In a first step, we explored whether the interaction context
short, lasting a few seconds, and reiterated after pauses. For had an influence on the sentence types used by owners when
the purpose of our analyses, we extracted all fragments of talking to their dog. The entire original corpus was examined,
speech directed to the dog and chained them in order to and utterances were categorized following the sentence types
123
Anim Cogn (2017) 20:499–509 503
used by Syrdal and Kim (2008): imperative, interrogative, conditions, repetitions of the same word by the same owner
affective and declarative (assertive), considered as four broad in the same condition were discarded.
modes of enunciation reflecting the speaker’s cognitive atti-
tude to the content. The beginning and end of sentences were Statistical analysis
determined on the basis of intonation and pauses. We distin-
guished four types of sentences on the basis of intonation, Acoustic analysis
meaning of the words/sentences and their context: (1)
imperative sentence type (i.e. ‘‘Sit!’’). We also looked for One-way repeated measure ANOVAs were performed to
imperatives disguised as questions (i.e. ‘‘Tu me donnes la test the influence of the different conditions (separation,
balle?’’, English translation: ‘‘You give me the ball?’’) that reunion, play and commands) on the owner PDS features;
were not ‘‘a request for information, but a directive’’ (Mitchell the sample size was n = 34.
and Edmonson 1999). (2) Interrogative sentence type were In addition, a two-way repeated measure ANOVA was
requests for information (i.e. ‘‘Quel jouet tu veux?’’, English performed to test the interaction between the variables
translation: ‘‘Which toy do you want?’’. (3) Affective sentence ‘‘condition’’ and ‘‘parental status’’. Post hoc analyses were
type included mainly positive exclamations: stock phrases and performed using Bonferroni-corrected paired t test (Sokal
common expressions (i.e. ‘‘A tout à l’heure!’’, ‘‘Bonjour/ and Rohlf 1995). Mann–Whitney tests were performed to
Coucou’’, English translation: ‘‘See you!’’ or ‘‘Hi!’’), inter- compare data from mothers versus non-mothers. The
jections (i.e. ‘‘hey!’’, ‘‘oh!’’, ‘‘allez!’’, ‘‘hop!’’) compliments sample size for these analyses was n = 33, because
(i.e. ‘‘t’es belle/t’es beau!’’, English translation: ‘‘You pretty information about the parental status of one participant was
girl/boy!’’) or verbal praising/congratulations (i.e. ‘‘bravo !’’, missing. Results are reported as mean ± SE (standard
‘‘bon chien !’’, English translation: ‘‘Good girl/boy!’’). Tag error). Two-tailed tests were used throughout.
questions (i.e. ‘‘C’est quoi ça?!’’, English translation: ‘‘What’s
this?’’) and post-completers (‘‘Hein?’’, English translation: Verbal analysis
‘‘You do?’’) (Mitchell and Edmonson 1999) were classed as
affective sentence type as the owner was not expecting an Because the data did not follow statistical normality (Sha-
answer but was trying to stimulate the dog. (4) Declarative piro–Wilk test for normality p \ 0.05) and could not be
sentence type was usually used by owners when talking in normalized with standard transformations, Friedman repe-
place of the dog or expressing the dog’s feelings (i.e. ‘‘Hmm, ated measures ANOVAs on ranks were performed to test the
je sens une odeur qui n’est pas la mienne’’, English translation: influence of the condition on the use of different sentence
‘‘Hmm, I am sensing a smell that’s not mine’’). Moreover, we types and particular utterances (Siegel and Castellan 1988).
measured the occurrence of particular utterances that did not Post hoc analyses were performed using Tukey’s HSD,
meet the precedent sentence types criteria: (5) attention-get- accounting for multiple comparisons and maintaining
ting devices (AGD) were considered, as calling the dog’s experiment-wise alpha at the specified level (0.05) (Maxwell
name and non-verbal sounds used to catch the dog’s attention and Delaney 2004).
(i.e. ‘‘qqq’’ or whistle). (6) Affectionate nicknames given to Moreover, Chi-square tests were performed to explore
the dog were the last (i.e. ‘‘ma douce’’, English translation: the distribution of the grammatical categories of words
‘‘sweety’’). used in each condition. Two-tailed tests were used
In a second step, sentences were fractionated into throughout. All statistical analyses were performed using
words and the total of different words used by owners SigmaPlot 13.0 software.
was extracted in each of the conditions. We first inves-
tigated words’ repetitions by calculating the total number
of words used with and without individual repetitions Results
and we performed the ratio between the two (see
‘‘Results’’). Acoustic analyses
Then, in order to explore whether the interaction context
had an influence on the grammatical categories of words Mean F0
used by owners, we selected words that were only used at
least by five participants in each condition and we classi- There was a significant effect of the condition on mean F0
fied the words into grammatical categories: nouns, pro- (one-way RM ANOVA, F(3, 33) = 2.92, p = 0.038).
nouns, verbs, adverbs, interjections, prepositions, Women used a higher-pitched voice in the reunion condi-
conjunctions, or articles. In this analysis, because some tion compared to the separation condition: Bonferroni
owners had a tendency to repeat the same words in paired t test = 2.88, p = 0.029.
123
504 Anim Cogn (2017) 20:499–509
There was a significant effect of the condition on the F0CV The data analysis revealed a significant main effect of the
[F(3, 33) = 12.05, p \ 0.001]. Women had a lower F0CV in owners’ parental status and of the condition of interaction on
the separation condition compared to all conditions: sepa- mean F0; no significant interaction effect was found, (two-
ration versus reunion: t = 5.58, p \ 0.001, separation way RM ANOVA, parental status effect, F(1, 31) = 4.33,
versus play: t = 4.60, p \ 0.001, separation versus com- p = 0.046; condition effect, F(3, 31) = 3.03, p = 0.033;
mands: t = 4.08, p \ 0.001. interaction, F(3, 31) = 0.10, p = 0.962). Non-mothers spoke
with a higher-pitched voice than mothers, with, respectively,
Coefficient of variation of the intensity contour (intCV) M = 305.6 ± SE = 7.8 and M = 278.5 ± SE = 10.4). No
significant effect of the parental status was found on other
There was a significant effect of the condition on the IntCV acoustic parameters.
[F(3, 33) = 6.05, p \ 0.001]. Women showed a greater A Mann–Whitney comparison test showed that non-
IntCV in the command condition compared to play: mothers (n = 21, M = 28.71 ± SE = 2.2) were also
t = 3.90, p = 0.001 and compared to reunion: t = 2.99, younger than mothers (n = 12, M = 46.0 ± SE = 2.4):
p = 0.021 (Fig. 2). U = 30, p \ 0.001. As such, we cannot rule out that the effect
of parental status on mean F0 was affected by the age of the
(a)
320
* speaker. We attempted to clarify this question by analysing
female owners’ speech directed to the experimenter (adult-
310 directed speech). This was possible because owners still wore
300 the lapel microphone during the separation condition.
A Mann–Whitney comparison test did not show any differ-
290
ence between non-mothers and mothers on mean F0 in adult-
280 directed speech (U = 86, p = 0.139, with, respectively,
270 M = 246.45 ± SE = 7.6 and M = 227.23 ± SE = 11.7).
260
Separation Reunion Play Commands Verbal analyses
123
Anim Cogn (2017) 20:499–509 505
sentences in the separation condition compared to all the reunion condition compared to the play condition
conditions: separation versus commands: q = 7.9, (q = 3.65, p \ 0.05) (Fig. 3).
p \ 0.05; separation versus reunion: q = 7.37, p \ 0.05;
separation versus play: q = 5.98, p \ 0.05. Repetitions
Declarative There was no significant effect of the con- Descriptive data from Table 1 showed that the ratio of the
dition on the use of declarative sentence type [v23 number of words with individual repetitions (IR) divided
(n = 34) = 4.54, p = 0.209] (Fig. 3). by the number of word without individual repetitions (IR)
Concerning particular utterances, there was a significant is close to 3. Overall, owners used a lot of repetitions. This
effect of the condition on the use of attention-getting ratio is particularly important in the commands condition
devices (AGD) [v23 (n = 34) = 24.66, p \ 0.001). Owners (2.85) and in the play condition (2.29). Repetitions are less
used significantly more AGD in commands and play con- important in the reunion condition, with a ratio of 1.86.
ditions compared to the other conditions: commands versus
separation: q = 3.72, p \ 0.05; commands versus reunion: Grammatical categories of words
q = 5.71, p \ 0.05; play versus reunion: q = 4.52,
p \ 0.05. There was also a significant effect of the con- Chi-square showed significant differences between condi-
dition on the use of nicknames [v23 (n = 34) = 18.89, tions with the use of grammatical categories (Fig. 4). In
p = 0.003]. Owners used significantly more nicknames in separation [v23 (n = 34) = 30.04, p \ 0.001), play [v23
(c) (d)
10 1,6
b
b b 1,4
8
1,2
6 1
0,8
4 0,6
a
2
0,4
0,2
0 0
Separation Reunion Play Commands Separation Reunion Play Commands
(e) (f)
5 c 2 b
4 b,c a,b
1,5
3
1
a,b
a
2 a,b
a 0,5
1
0 0
Separation Reunion Play Commands Separation Reunion Play Commands
123
506 Anim Cogn (2017) 20:499–509
(n = 34) = 37.64, p \ 0.001) and commands [v23 show that women adapted both the verbal and non-verbal
(n = 34) = 62.05, p \ 0.001), data showed that partici- aspects of their utterances to the interaction context.
pants used principally verbs when addressing the dog. In the separation condition, female dog owners used a
However, in the reunion condition [v23 (n = 34) = 8.04, low-pitched voice, flat intonation contours and high voice
p = 0.09), participants did not use a particular grammatical intensity variation. Additionally, they used verbal features
category, but multiple ones (interjections, adverbs, verbs consistent with these acoustic parameters: a high frequency
and other). of imperatives and few affective sentences. With regard to
our initial hypotheses, these results support the idea that
women tried to prevent the dog to come with them using an
Discussion imperative style of communication rather than a soothing
style of interaction. According to our predictions, in the
To our knowledge, the present study is the first to inves- reunion condition women used a high-pitched voice and
tigate both the verbal and non-verbal features of adult significantly more pitch variations. These non-verbal fea-
women’s speech during different types of interactions with tures were reinforced by the use of affective sentences and
their dogs. Speech utterances were recorded in four affectionate nicknames. Moreover, they did not use verbs
experimental conditions which solicit different emotions or imperatives as frequently as in the other types of inter-
and intents from the owners: before a brief separation, after actions. Thus, our findings suggest that female dog owners
reunion, during play and giving commands. Our results use a high-pitched voice when they want to express praise
123
Anim Cogn (2017) 20:499–509 507
or affection and a low-pitched voice when they want to Ohala (1984), the F0 of voice is used as a gesture which
control the dog. This result confirms previous studies accompanies or is superimposed onto the linguistic mes-
indicating that increases in mean F0 are typical of vocal sage, in order to enhance its meaning. We suggest that
expressions of enjoyment, happiness, whereas decreases in women use specific linguistic and verbal patterns when
mean F0 characterize expressions of authority, irritation speaking to their dog because they aim to teach or reinforce
and anger (Bolinger 1964; Fernald 1992). Moreover, it is basic commands. Indeed, owners report that they feel that
consistent with ethological observations highlighted by their dog understands the meaning of certain verbal utter-
Morton (1977) that non-verbal auditory signals, in humans ances (Pongrácz et al. 2001). Thus, when women speak to
and non-humans, have similar functions: high tonal sounds their dogs, both the prosody and the verbal content appear
in animals repertoires are associated with affiliative or to function to convent context-relevant information. Fur-
submissive motivation while low, harsh sounds are asso- ther studies should investigate whether dogs whose owners
ciated with aggressive motivation. Hence, because dogs do frequently use PDS are more capable of acquiring and
not possess the ability of language, female dog owners understanding commands than dogs whose owners seldom
emphasize non-verbal cues to convey their emotions and communicate using PDS.
intents, using signal components with cross-specific value. We also hypothesized that PDS would be influenced by
In the play condition, in line with our predictions, the parental status of the owner, because non-parents tend
women frequently used attention-getting devices (AGD), to see their dog as a family member significantly more than
questions and imperatives although we noticed that parents. We predicted that non-mothers would present
imperatives were more encouragements than real com- exacerbated prosodic features compared to mothers. Our
mands. We found a similar pattern of verbal features in the results are consistent with this hypothesis: non-mothers
commands condition: women used a high frequency of spoke to their dog with a higher-pitched voice compared to
AGD and imperatives; contrary to the play condition, they mothers. In our sample, non-mothers were younger than
did not use questions as they were not engaged in a mutual mothers which may account for this result, as some authors
activity and were just asking the dog to obey. In addition, report a decrease in voice pitch with age in adult women
women did not use a significantly low-pitched voice or flat (Harrington et al. 2007). However, we failed to find a
contours. A possible interpretation is that when female significant difference between non-mothers and mothers’
owners give a command to their dog, they first attempt to fundamental frequency mean in ADS in our sample. One
draw his/her attention, using exaggerated acoustic features explanation for these results could be that non-mothers
and then seek to maintain the dogs’ arousal and motivation have a stronger tendency to express friendliness and
throughout the activity. Female owners used prosody as an affection to their dogs, consistent with research indicating
ostensive cue, i.e. as a means to alert the dog to the com- that people who do not live with children tend to be more
munication directed to them (Topál et al. 2014). Acoustic attached to their dogs (Marinelli et al. 2007) and that
features characterized by decreasing F0 are better related to couples without children show a particularly high degree of
prohibition (Fernald 1989; Trainor et al. 2000), as observed attachment to their pet and are more prone to consider their
in the separation condition when female owners intend to dog as a member of their family, even as their own child
inhibit the dog’s behaviour. (Berryman et al. 1984; Albert and Bulcroft 1988; Kidd and
The current study thus confirms that women do not rely Kidd 1989; Poresky and Daniels 1998; Taylor 2006; Del
solely on prosody when they communicate with dogs as Monte Foods 2011).
they associate specific verbal content to the non-verbal
component of their vocal signals. We observed that female
owners spoke to their dog using words and constructions Conclusion
derived from normal speech (e.g. similar to those used in
ADS), but with a limited vocabulary including frequent Overall, our results confirm that women speaking to their
repetitions of isolated words or phrases. They also sim- dog, like mothers speaking to their babies, adapt the verbal
plified their utterances using verbs but excluding gram- and non-verbal features of their speech to the interaction’s
matical categories that are likely to be meaningless to the context in order to provide the dog with information about
dog, such as articles. Similar linguistic patterns have been their intentions and emotions. They also take into account
identified in IDS and are believed to facilitate infant’ lan- the dog’s limited ability to understand human verbal sig-
guage comprehension and acquisition (Saint-Georges et al. nals and adjust their mix of speech components accord-
2013). When interacting with their dog, female owners ingly, using appropriate verbal content reinforced by non-
may use acoustic features such as pitch modulations as a verbal cues in order to enhance the dog’s comprehension
tool to highlight focal words and thus to assist words’ and to some extent with the aim of teaching the dog some
segmentation and recognition by dogs. As mentioned by basics utterances.
123
508 Anim Cogn (2017) 20:499–509
While research has shown that naturally spoken IDS is Burnham D et al (1998) Are you my little pussy-cat? Acoustic,
associated with increased infant attention and social phonetic and affective qualities of infant-and pet-directed
speech. In: Fifth international conference on spoken language
responsiveness compared to ADS (Dunst et al. 2012) and processing (ICSLP)
that IDS leads to an increase neuronal activation (Golinkoff Burnham D, Kitamura C, Vollmer-Conna U (2002) What’s new,
et al. 2015) as well as to positive effects on infant language pussycat? On talking to babies and animals. Science 296:1435.
development (Liu et al. 2003; Song et al. 2010), no study doi:10.1126/science.1069587
Del Monte Foods, Business Wire (2011) New study reveals that the
has yet explored dogs’ preference for PDS. Future studies American family has gone to the dogs. The Milo’s KitchenTMPet
should investigate the perceptual relevance of PDS’s fea- Parent Survey, conducted by Kelton Research. http://www.
tures from dogs’ perspective, in order to evaluate the effect businesswire.com/news/home/20110502006312/en/Study-
of PDS on dogs’ attention, mood and ability to learn new Reveals-American-Family-Dogs. Accessed 2 May 2011
Dunst C, Gorman E, Hamby D (2012) Preference for infant-directed
commands. speech in preverbal young children. Cent Early Lit Learn Rev
5:1–13
Acknowledgements The authors are indebted to David Reby for his Fernald A (1989) Intonation and communicative intent in mothers’
invaluable contribution to this article. We would like to thank Dr. speech to infants: is the melody the message? Child Dev
S. Perrot (IRCA-ENVA) for providing access to the IRCA room and 60:1497–1510. doi:10.2307/1130938
its facilities at the ENVA. Thanks to Marine Parker, Raphaëlle Fernald A (1992) Meaningful melodies in mothers’ speech to infants.
Bourrec, Raphaëlle Tigeot and Justine Guillaumont for their partici- In: Papoušek H, Papoušek M (eds) Nonverbal vocal communi-
pation in these experiments, as well as for the pilot experiment. cation: comparative and developmental approaches. Cambridge
Thanks to Mathieu Amy for help in statistical analyses. Thanks to the University Press, Cambridge, pp 262–282
CHUVA (ENVA) for help with the recruitment of owners. Thanks to Fernald A, Simon T (1984) Expanded intonation contours in mothers’
owners who accepted to take part to this study. speech to newborns. Dev Psychol 20:104–113. doi:10.1037/
0012-1649.20.1.104
Compliance with ethical standards Fogle B (1992) The dog’s mind: understanding your dog’s behavior.
Wiley, New York
Conflict of interest The authors declare that they have no conflict of Gibson JM, Scavelli SA, Udell CJ, Udell MAR (2014) Domestic dogs
interest. (Canis lupus familiaris) are sensitive to the ‘‘human’’ qualities of
vocal commands. Anim Behav Cogn 1:281–295. doi:10.12966/
Ethical approval All procedures performed in studies involving abc.08.05.2014
human participants were in accordance with the ethical standards of Golinkoff RM, Can DD, Soderstrom M, Hirsh-Pasek K (2015) (Baby)
the institutional and/or national research committee and with the 1964 talk to me: the social context of infant-directed speech and its
Helsinki Declaration and its later amendments or comparable ethical effects on early language acquisition. Curr Dir Psychol Sci
standards. All applicable international, national and/or institutional 24:339–344. doi:10.1177/0963721415595345
guidelines for the care and use of animals were followed and ‘‘all Harrington J, Palethorpe S, Watson CI (2007) Age-related changes in
procedures performed in studies involving animals were in accor- fundamental frequency and formants: a longitudinal study of
dance with the ethical standards of the institution at which the study four speakers. In: Interspeech 2007: 8th annual conference of the
was conducted’’. The study received the approval of the ethical international speech communication association, vol 2,
committee of ENVA (COMERC), no. 2015-03-11. pp 1081–1084
Hirsh-Pasek K, Treiman R (1982) Doggerel: Motherese in a new context.
J Child Lang 9:229–237. doi:10.1017/S0305000900003731
Horowitz A, Hecht J (2016) Examining dog-human play: the
References characteristics, affect, and vocalizations of a unique interspecific
interaction. Anim Cogn 19:779–788. doi:10.1007/s10071-016-
Ainsworth MDS, Wittig BA (1969) Attachment and exploratory 0976-3
behavior of one-year-olds in a strange situation. In: Foss BM Kidd AH, Kidd RM (1989) Factors in adults’ attitudes toward pets.
(ed) Determinants of infant behavior, vol 4. Methuen, London, Psychol Rep 65:903–910. doi:10.2466/pr0.1990.66.3.775
pp 111–136 Kim H, Diehl MM, Panneton R, Moon C (2006) Hyperarticulation in
Albert A, Bulcroft K (1988) Pets, families, and the life course. mothers’ speech to babies and puppies. Paper presented at the
J Marriage Fam 50:543–552. doi:10.2307/352019 annual meeting of the XVth biennial international conference on
Archer J (1997) Why do people love their pets? Evol Hum Behav infant studies, Westin Miyako, Kyoto, Japan. http://citation.
18:237–259. doi:10.1016/S0162-3095(99)80001-4 allacademic.com/meta/p94616_index.html
Archer J, Monton S (2011) Preferences for infant facial features in pet Liu HM, Kuhl PK, Tsao FM (2003) An association between mothers’
dogs and cats. Ethology 117:217–226. doi:10.1111/j.1439-0310. speech clarity and infants’ speech discrimination skills. Dev Sci
2010.01863.x 6:F1–F10. doi:10.1111/1467-7687.00275
Askew HR (1996) Treatment of behavior problems in dogs and cats: a Marinelli L, Adamelli S, Normando S, Bono G (2007) Quality of life
guide for the small animal veterinarian. Wiley, London of the pet dog: Influence of owner and dog’s characteristics.
Berryman JC, Howells K, Lloyd-Evans M (1984) Pet owner attitudes Appl Anim Behav Sci 108:143–156. doi:10.1016/j.applanim.
to pets and people: a psychological study. Vet Rec 117:659–661. 2006.11.018
doi:10.1136/vr.117.25-26.659 Markwell PJ, Thorne CJ (1987) Early behavioural development of
Bolinger DL (1964) Around the age of language: Intonation. In: dogs. J Small Anim Pract 28:984–991. doi:10.1111/j.1748-5827.
Bolinger D (ed) Intonation. Penguin Books, Harmondsworth, 1987.tb01322.x
pp 19–29 Maxwell SE, Delaney HD (2004) Designing experiments and
Bowlby J (1969) Attachement et perte, vol 1: L’attachement. PUF, analyzing data: a model comparison perspective, 2nd edn.
Paris Psychology Press, New York
123
Anim Cogn (2017) 20:499–509 509
Mitchell RW (2001) Americans’ talk to dogs: similarities and Prato-Previde E, Fallani G, Valsecchi P (2006) Gender differences in
differences with talk to infants. Res Lang Soc Interact owners interacting with pet dogs: an observational study.
34:183–210. doi:10.1207/S15327973RLSI34-2_2 Ethology 112:64–73. doi:10.1111/j.1439-0310.2006.01123.x
Mitchell RW (2004) Controlling the dog, pretending to have a Rehn T, Handlin L, Uvnäs-Moberg K, Keeling LJ (2013) Dogs’
conversation, or just being friendly? Influences of sex and endocrine and behavioural responses at reunion are affected by
familiarity on Americans’ talk to dogs during play. Interact Stud how the human initiates contact. Physiol Behav 124:45–53.
5:99–129. doi:10.1075/is.5.1.06mit doi:10.1016/j.physbeh.2013.10.009
Mitchell RW, Edmonson E (1999) Functions of repetitive talk to dogs Rogers J, Hart LA, Boltz RP (1993) The role of pet dogs in casual
during play: control, conversation, or planning? Soc Anim conversations of elderly adults. J Soc Psychol 133:265–277.
7:55–81. doi:10.1163/156853099X00167 doi:10.1080/00224545.1993.9712145
Morton ES (1977) On the occurrence and significance of motivation- Saint-Georges C, Chetouani M, Cassel R, Apicella F, Mahdhaoui A,
structural rules in some bird and mammal sounds. Am Nat Muratori F et al (2013) Motherese in interaction: at the cross-
111:855–869. doi:10.1086/283219 road of emotion and cognition? (A systematic review). PLoS
Nagasawa M, Kikusui T, Onaka T, Ohta M (2009) Dog’s gaze at its ONE 8:e78103. doi:10.1371/journal.pone.0078103
owner increases owner’s urinary oxytocin during social interac- Siegel S, Castellan NJ (1988) Book review: nonparametric statistics
tion. Horm Behav 55:434–441. doi:10.1016/j.yhbeh.2008.12.002 for the behavioral Sciences, 2nd edn. McGraw-Hill, New York
Nagasawa M, Mitsui S, En S, Ohtani N, Ohta M, Sakuma Y et al Sokal RR, Rohlf FJ (1995) Biometry: the principals and practice of
(2015) Oxytocin-gaze positive loop and the coevolution of statistics in biological research, 3rd edn. WH Freman and
human–dog bonds. Science 348:333–336. doi:10.1126/science. Company, New York
1261022 Sokol RI, Webster KL, Thompson NS, Stevens DA (2005) Whining
Newport E, Gleitman H, Gleitman L (1977) Mother, I’d rather do it as mother-directed speech. Infant Child Dev 14:478–490. doi:10.
myself: some effects and non-effects of maternal speech style. 1002/icd.420
In: Snow CE, Ferguson CA (eds) Talking to children. Cambridge Song JY, Demuth K, Morgan JL (2010) Effects of the acoustic
University Press, Cambridge, pp 109–149 properties of infant-directed speech on infant word recognition.
Odendaal JSJ, Meintjes RA (2003) Neurophysiological correlates of J Acoust Soc Am 128:352–363. doi:10.1121/1.3419786
affiliative behaviour between humans and dogs. Vet J Lond Engl Syrdal A, Kim YJ (2008) Dialog speech acts and prosody: consid-
165:296–301. doi:10.1016/S1090-0233(02)00237-X erations for TTS. In: Proceedings of speech prosody. Campinas,
Ohala JJ (1984) An ethological perspective on common cross- Brazil, pp 661–665
language utilization of F0 of voice. Phonetica 41:1–16. doi:10. Taylor P (2006) Gauging family intimacy: Pew Research Centers
1159/000261706 Social Demographic Trends Project RSS. Pew Research Center
Palestrini C, Previde EP, Spiezio C, Verga M (2005) Heart rate and http://www.pewsocialtrends.org/2006/03/07/gauging-family-
behavioural responses of dogs in the Ainsworth’s Strange intimacy/
Situation: a pilot study. Appl Anim Behav Sci 94:75–88. Thalmann O, Shapiro B, Cui P, Schuenemann VJ, Sawyer SK,
doi:10.1016/j.applanim.2005.02.005 Greenfield DL et al (2013) Complete mitochondrial genomes of
Papoušek M, Bornstein MH, Nuzzo C, Papoušek H, Symmes D ancient canids suggest a European origin of domestic dogs.
(1990) Infant responses to prototypical melodic contours in Science 342:871–874. doi:10.1126/science.1243650
parental speech. Inf Behav Dev 13:539–545. doi:10.1016/0163- Topál J, Miklósi Á, Csányi V, Dóka A (1998) Attachment behavior in
6383(90)90022-Z dogs (Canis familiaris): a new application of Ainsworth’s (1969)
Papoušek M, Papoušek H, Symmes D (1991) The meanings of Strange Situation Test. J Comp Psychol 112:219. doi:10.1037/
melodies in motherese in tone and stress languages. Infant Behav 0735-7036.112.3.219
Dev 14:415–440. doi:10.1016/0163-6383(91)90031-M Topál J, Kis A, Oláh K (2014) Dogs’ sensitivity to human ostensive
Pongrácz P, Miklósi A, Csányi V (2001) Owner’s beliefs on the cues: a unique adaptation. In: The social dog. Elsevier, London,
ability of their pet dogs to understand human verbal communi- pp 319–346
cation: a case of social understanding. Curr Psychol Cogn Trainor LJ, Clark ED, Huntley A, Adams BA (1997) The acoustic
20:87–107 basis of preferences for infant-directed singing. Infant Behav
Poresky RH, Daniels AM (1998) Demographics of pet presence and Dev 20:383–396. doi:10.1016/S0163-6383(97)90009-6
attachment. Anthrozoös 11:236–241. doi:10.2752/ Trainor LJ, Austin CM, Desjardins RN (2000) Is infant-directed
089279398787000508 speech prosody a result of the vocal expression of emotion?
Prato-Previde E, Custance DM, Spiezio C, Sabatini F (2003) Is the Psychol Sci 11:188–195. doi:10.1111/1467-9280.00240
dog-human relationship an attachment bond? An observational Xu N, Burnham D, Kitamura C, Vollmer-Conna U (2013) Vowel
study using Ainsworth’s strange situation. Behaviour hyperarticulation in parrot-, dog-and infant-directed speech. Anthro-
140:225–254. doi:10.1163/156853903321671514 zoos 26:373–380. doi:10.2752/175303713X13697429463592
123